[PATCH AUTOSEL 5.4 014/187] netfilter: ctnetlink: netns exit must wait for callbacks

[PATCH AUTOSEL 5.4 014/187] netfilter: ctnetlink: netns exit must wait for callbacks

[Sasha Levin]

From: Florian Westphal <fw@strlen.de>

[ Upstream commit 18a110b022a5c02e7dc9f6109d0bd93e58ac6ebb ]

Curtis Taylor and Jon Maxwell reported and debugged a crash on 3.10
based kernel.

Crash occurs in ctnetlink_conntrack_events because net->nfnl socket is
NULL.  The nfnl socket was set to NULL by netns destruction running on
another cpu.

The exiting network namespace calls the relevant destructors in the
following order:

1. ctnetlink_net_exit_batch

This nulls out the event callback pointer in struct netns.

2. nfnetlink_net_exit_batch

This nulls net->nfnl socket and frees it.

3. nf_conntrack_cleanup_net_list

This removes all remaining conntrack entries.

This is order is correct. The only explanation for the crash so ar is:

cpu1: conntrack is dying, eviction occurs:
 -> nf_ct_delete()
   -> nf_conntrack_event_report \
     -> nf_conntrack_eventmask_report
       -> notify->fcn() (== ctnetlink_conntrack_events).

cpu1: a. fetches rcu protected pointer to obtain ctnetlink event callback.
      b. gets interrupted.
 cpu2: runs netns exit handlers:
     a runs ctnetlink destructor, event cb pointer set to NULL.
     b runs nfnetlink destructor, nfnl socket is closed and set to NULL.
cpu1: c. resumes and trips over NULL net->nfnl.

Problem appears to be that ctnetlink_net_exit_batch only prevents future
callers of nf_conntrack_eventmask_report() from obtaining the callback.
It doesn't wait of other cpus that might have already obtained the
callbacks address.

I don't see anything in upstream kernels that would prevent similar
crash: We need to wait for all cpus to have exited the event callback.

Fixes: 9592a5c01e79dbc59eb56fa ("netfilter: ctnetlink: netns support")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/netfilter/nf_conntrack_netlink.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c
index e2d13cd18875..aa8adf930b3c 100644
--- a/net/netfilter/nf_conntrack_netlink.c
+++ b/net/netfilter/nf_conntrack_netlink.c
@@ -3602,6 +3602,9 @@ static void __net_exit ctnetlink_net_exit_batch(struct list_head *net_exit_list)
 
 	list_for_each_entry(net, net_exit_list, exit_list)
 		ctnetlink_net_exit(net);
+
+	/* wait for other cpus until they are done with ctnl_notifiers */
+	synchronize_rcu();
 }
 
 static struct pernet_operations ctnetlink_net_ops = {
-- 
2.20.1

[PATCH AUTOSEL 5.4 017/187] netfilter: nf_tables_offload: Check for the NETDEV_UNREGISTER event

[Sasha Levin]

From: wenxu <wenxu@ucloud.cn>

[ Upstream commit d1f4c966475c6dd2545c6625022cb24e878bee11 ]

Check for the NETDEV_UNREGISTER event from the nft_offload_netdev_event
function, which is the event that actually triggers the clean up.

Fixes: 06d392cbe3db ("netfilter: nf_tables_offload: remove rules when the device unregisters")
Signed-off-by: wenxu <wenxu@ucloud.cn>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/netfilter/nf_tables_offload.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/netfilter/nf_tables_offload.c b/net/netfilter/nf_tables_offload.c
index 5f6037695dee..6f7eab502e65 100644
--- a/net/netfilter/nf_tables_offload.c
+++ b/net/netfilter/nf_tables_offload.c
@@ -446,6 +446,9 @@ static int nft_offload_netdev_event(struct notifier_block *this,
 	struct net *net = dev_net(dev);
 	struct nft_chain *chain;
 
+	if (event != NETDEV_UNREGISTER)
+		return NOTIFY_DONE;
+
 	mutex_lock(&net->nft.commit_mutex);
 	chain = __nft_offload_get_chain(dev);
 	if (chain)
-- 
2.20.1

[PATCH AUTOSEL 5.4 025/187] netfilter: nf_queue: enqueue skbs with NULL dst

[Sasha Levin]

From: Marco Oliverio <marco.oliverio@tanaza.com>

[ Upstream commit 0b9173f4688dfa7c5d723426be1d979c24ce3d51 ]

Bridge packets that are forwarded have skb->dst == NULL and get
dropped by the check introduced by
b60a77386b1d4868f72f6353d35dabe5fbe981f2 (net: make skb_dst_force
return true when dst is refcounted).

To fix this we check skb_dst() before skb_dst_force(), so we don't
drop skb packet with dst == NULL. This holds also for skb at the
PRE_ROUTING hook so we remove the second check.

Fixes: b60a77386b1d ("net: make skb_dst_force return true when dst is refcounted")
Signed-off-by: Marco Oliverio <marco.oliverio@tanaza.com>
Signed-off-by: Rocco Folino <rocco.folino@tanaza.com>
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/netfilter/nf_queue.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/netfilter/nf_queue.c b/net/netfilter/nf_queue.c
index a2b58de82600..f8f52ff99cfb 100644
--- a/net/netfilter/nf_queue.c
+++ b/net/netfilter/nf_queue.c
@@ -189,7 +189,7 @@ static int __nf_queue(struct sk_buff *skb, const struct nf_hook_state *state,
 		goto err;
 	}
 
-	if (!skb_dst_force(skb) && state->hook != NF_INET_PRE_ROUTING) {
+	if (skb_dst(skb) && !skb_dst_force(skb)) {
 		status = -ENETDOWN;
 		goto err;
 	}
-- 
2.20.1

[PATCH AUTOSEL 5.4 032/187] netfilter: uapi: Avoid undefined left-shift in xt_sctp.h

[Sasha Levin]

From: Phil Sutter <phil@nwl.cc>

[ Upstream commit 164166558aacea01b99c8c8ffb710d930405ba69 ]

With 'bytes(__u32)' being 32, a left-shift of 31 may happen which is
undefined for the signed 32-bit value 1. Avoid this by declaring 1 as
unsigned.

Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 include/uapi/linux/netfilter/xt_sctp.h | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/include/uapi/linux/netfilter/xt_sctp.h b/include/uapi/linux/netfilter/xt_sctp.h
index 4bc6d1a08781..b4d804a9fccb 100644
--- a/include/uapi/linux/netfilter/xt_sctp.h
+++ b/include/uapi/linux/netfilter/xt_sctp.h
@@ -41,19 +41,19 @@ struct xt_sctp_info {
 #define SCTP_CHUNKMAP_SET(chunkmap, type) 		\
 	do { 						\
 		(chunkmap)[type / bytes(__u32)] |= 	\
-			1 << (type % bytes(__u32));	\
+			1u << (type % bytes(__u32));	\
 	} while (0)
 
 #define SCTP_CHUNKMAP_CLEAR(chunkmap, type)		 	\
 	do {							\
 		(chunkmap)[type / bytes(__u32)] &= 		\
-			~(1 << (type % bytes(__u32)));	\
+			~(1u << (type % bytes(__u32)));	\
 	} while (0)
 
 #define SCTP_CHUNKMAP_IS_SET(chunkmap, type) 			\
 ({								\
 	((chunkmap)[type / bytes (__u32)] & 		\
-		(1 << (type % bytes (__u32)))) ? 1: 0;	\
+		(1u << (type % bytes (__u32)))) ? 1: 0;	\
 })
 
 #define SCTP_CHUNKMAP_RESET(chunkmap) \
-- 
2.20.1

[PATCH AUTOSEL 5.4 033/187] netfilter: nft_set_rbtree: bogus lookup/get on consecutive elements in named sets

[Sasha Levin]

From: Pablo Neira Ayuso <pablo@netfilter.org>

[ Upstream commit db3b665dd77b34e34df00e17d7b299c98fcfb2c5 ]

The existing rbtree implementation might store consecutive elements
where the closing element and the opening element might overlap, eg.

	[ a, a+1) [ a+1, a+2)

This patch removes the optimization for non-anonymous sets in the exact
matching case, where it is assumed to stop searching in case that the
closing element is found. Instead, invalidate candidate interval and
keep looking further in the tree.

The lookup/get operation might return false, while there is an element
in the rbtree. Moreover, the get operation returns true as if a+2 would
be in the tree. This happens with named sets after several set updates.

The existing lookup optimization (that only works for the anonymous
sets) might not reach the opening [ a+1,... element if the closing
...,a+1) is found in first place when walking over the rbtree. Hence,
walking the full tree in that case is needed.

This patch fixes the lookup and get operations.

Fixes: e701001e7cbe ("netfilter: nft_rbtree: allow adjacent intervals with dynamic updates")
Fixes: ba0e4d9917b4 ("netfilter: nf_tables: get set elements via netlink")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/netfilter/nft_set_rbtree.c | 21 ++++++++++++++++-----
 1 file changed, 16 insertions(+), 5 deletions(-)

diff --git a/net/netfilter/nft_set_rbtree.c b/net/netfilter/nft_set_rbtree.c
index 57123259452f..a9f804f7a04a 100644
--- a/net/netfilter/nft_set_rbtree.c
+++ b/net/netfilter/nft_set_rbtree.c
@@ -74,8 +74,13 @@ static bool __nft_rbtree_lookup(const struct net *net, const struct nft_set *set
 				parent = rcu_dereference_raw(parent->rb_left);
 				continue;
 			}
-			if (nft_rbtree_interval_end(rbe))
-				goto out;
+			if (nft_rbtree_interval_end(rbe)) {
+				if (nft_set_is_anonymous(set))
+					return false;
+				parent = rcu_dereference_raw(parent->rb_left);
+				interval = NULL;
+				continue;
+			}
 
 			*ext = &rbe->ext;
 			return true;
@@ -88,7 +93,7 @@ static bool __nft_rbtree_lookup(const struct net *net, const struct nft_set *set
 		*ext = &interval->ext;
 		return true;
 	}
-out:
+
 	return false;
 }
 
@@ -139,8 +144,10 @@ static bool __nft_rbtree_get(const struct net *net, const struct nft_set *set,
 			if (flags & NFT_SET_ELEM_INTERVAL_END)
 				interval = rbe;
 		} else {
-			if (!nft_set_elem_active(&rbe->ext, genmask))
+			if (!nft_set_elem_active(&rbe->ext, genmask)) {
 				parent = rcu_dereference_raw(parent->rb_left);
+				continue;
+			}
 
 			if (!nft_set_ext_exists(&rbe->ext, NFT_SET_EXT_FLAGS) ||
 			    (*nft_set_ext_flags(&rbe->ext) & NFT_SET_ELEM_INTERVAL_END) ==
@@ -148,7 +155,11 @@ static bool __nft_rbtree_get(const struct net *net, const struct nft_set *set,
 				*elem = rbe;
 				return true;
 			}
-			return false;
+
+			if (nft_rbtree_interval_end(rbe))
+				interval = NULL;
+
+			parent = rcu_dereference_raw(parent->rb_left);
 		}
 	}
 
-- 
2.20.1

[PATCH AUTOSEL 5.4 034/187] netfilter: nf_tables: validate NFT_SET_ELEM_INTERVAL_END

[Sasha Levin]

From: Pablo Neira Ayuso <pablo@netfilter.org>

[ Upstream commit bffc124b6fe37d0ae9b428d104efb426403bb5c9 ]

Only NFTA_SET_ELEM_KEY and NFTA_SET_ELEM_FLAGS make sense for elements
whose NFT_SET_ELEM_INTERVAL_END flag is set on.

Fixes: 96518518cc41 ("netfilter: add nftables")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/netfilter/nf_tables_api.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index 712a428509ad..7120eba71ac5 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -4489,14 +4489,20 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set,
 		if (nla[NFTA_SET_ELEM_DATA] == NULL &&
 		    !(flags & NFT_SET_ELEM_INTERVAL_END))
 			return -EINVAL;
-		if (nla[NFTA_SET_ELEM_DATA] != NULL &&
-		    flags & NFT_SET_ELEM_INTERVAL_END)
-			return -EINVAL;
 	} else {
 		if (nla[NFTA_SET_ELEM_DATA] != NULL)
 			return -EINVAL;
 	}
 
+	if ((flags & NFT_SET_ELEM_INTERVAL_END) &&
+	     (nla[NFTA_SET_ELEM_DATA] ||
+	      nla[NFTA_SET_ELEM_OBJREF] ||
+	      nla[NFTA_SET_ELEM_TIMEOUT] ||
+	      nla[NFTA_SET_ELEM_EXPIRATION] ||
+	      nla[NFTA_SET_ELEM_USERDATA] ||
+	      nla[NFTA_SET_ELEM_EXPR]))
+		return -EINVAL;
+
 	timeout = 0;
 	if (nla[NFTA_SET_ELEM_TIMEOUT] != NULL) {
 		if (!(set->flags & NFT_SET_TIMEOUT))
-- 
2.20.1

[PATCH AUTOSEL 5.4 035/187] netfilter: nf_tables: validate NFT_DATA_VALUE after nft_data_init()

[Sasha Levin]

From: Pablo Neira Ayuso <pablo@netfilter.org>

[ Upstream commit 0d2c96af797ba149e559c5875c0151384ab6dd14 ]

Userspace might bogusly sent NFT_DATA_VERDICT in several netlink
attributes that assume NFT_DATA_VALUE. Moreover, make sure that error
path invokes nft_data_release() to decrement the reference count on the
chain object.

Fixes: 96518518cc41 ("netfilter: add nftables")
Fixes: 0f3cd9b36977 ("netfilter: nf_tables: add range expression")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/netfilter/nf_tables_api.c |  4 +++-
 net/netfilter/nft_bitwise.c   |  4 ++--
 net/netfilter/nft_cmp.c       |  6 ++++++
 net/netfilter/nft_range.c     | 10 ++++++++++
 4 files changed, 21 insertions(+), 3 deletions(-)

diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index 7120eba71ac5..4c03c14e46bc 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -4252,8 +4252,10 @@ static int nft_get_set_elem(struct nft_ctx *ctx, struct nft_set *set,
 		return err;
 
 	err = -EINVAL;
-	if (desc.type != NFT_DATA_VALUE || desc.len != set->klen)
+	if (desc.type != NFT_DATA_VALUE || desc.len != set->klen) {
+		nft_data_release(&elem.key.val, desc.type);
 		return err;
+	}
 
 	priv = set->ops->get(ctx->net, set, &elem, flags);
 	if (IS_ERR(priv))
diff --git a/net/netfilter/nft_bitwise.c b/net/netfilter/nft_bitwise.c
index 02afa752dd2e..10e9d50e4e19 100644
--- a/net/netfilter/nft_bitwise.c
+++ b/net/netfilter/nft_bitwise.c
@@ -80,7 +80,7 @@ static int nft_bitwise_init(const struct nft_ctx *ctx,
 			    tb[NFTA_BITWISE_MASK]);
 	if (err < 0)
 		return err;
-	if (d1.len != priv->len) {
+	if (d1.type != NFT_DATA_VALUE || d1.len != priv->len) {
 		err = -EINVAL;
 		goto err1;
 	}
@@ -89,7 +89,7 @@ static int nft_bitwise_init(const struct nft_ctx *ctx,
 			    tb[NFTA_BITWISE_XOR]);
 	if (err < 0)
 		goto err1;
-	if (d2.len != priv->len) {
+	if (d2.type != NFT_DATA_VALUE || d2.len != priv->len) {
 		err = -EINVAL;
 		goto err2;
 	}
diff --git a/net/netfilter/nft_cmp.c b/net/netfilter/nft_cmp.c
index 0744b2bb46da..ae730dba60c8 100644
--- a/net/netfilter/nft_cmp.c
+++ b/net/netfilter/nft_cmp.c
@@ -80,6 +80,12 @@ static int nft_cmp_init(const struct nft_ctx *ctx, const struct nft_expr *expr,
 	if (err < 0)
 		return err;
 
+	if (desc.type != NFT_DATA_VALUE) {
+		err = -EINVAL;
+		nft_data_release(&priv->data, desc.type);
+		return err;
+	}
+
 	priv->sreg = nft_parse_register(tb[NFTA_CMP_SREG]);
 	err = nft_validate_register_load(priv->sreg, desc.len);
 	if (err < 0)
diff --git a/net/netfilter/nft_range.c b/net/netfilter/nft_range.c
index 4701fa8a45e7..89efcc5a533d 100644
--- a/net/netfilter/nft_range.c
+++ b/net/netfilter/nft_range.c
@@ -66,11 +66,21 @@ static int nft_range_init(const struct nft_ctx *ctx, const struct nft_expr *expr
 	if (err < 0)
 		return err;
 
+	if (desc_from.type != NFT_DATA_VALUE) {
+		err = -EINVAL;
+		goto err1;
+	}
+
 	err = nft_data_init(NULL, &priv->data_to, sizeof(priv->data_to),
 			    &desc_to, tb[NFTA_RANGE_TO_DATA]);
 	if (err < 0)
 		goto err1;
 
+	if (desc_to.type != NFT_DATA_VALUE) {
+		err = -EINVAL;
+		goto err2;
+	}
+
 	if (desc_from.len != desc_to.len) {
 		err = -EINVAL;
 		goto err2;
-- 
2.20.1

[PATCH AUTOSEL 5.4 036/187] netfilter: nf_tables: skip module reference count bump on object updates

[Sasha Levin]

From: Pablo Neira Ayuso <pablo@netfilter.org>

[ Upstream commit fd57d0cbe187e93f63777d36e9f49293311d417f ]

Use __nft_obj_type_get() instead, otherwise there is a module reference
counter leak.

Fixes: d62d0ba97b58 ("netfilter: nf_tables: Introduce stateful object update operation")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/netfilter/nf_tables_api.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index 4c03c14e46bc..67ca47c7ce54 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -5217,7 +5217,7 @@ static int nf_tables_newobj(struct net *net, struct sock *nlsk,
 		if (nlh->nlmsg_flags & NLM_F_REPLACE)
 			return -EOPNOTSUPP;
 
-		type = nft_obj_type_get(net, objtype);
+		type = __nft_obj_type_get(objtype);
 		nft_ctx_init(&ctx, net, skb, nlh, family, table, NULL, nla);
 
 		return nf_tables_updobj(&ctx, type, nla[NFTA_OBJ_DATA], obj);
-- 
2.20.1

[PATCH AUTOSEL 5.4 037/187] netfilter: nf_tables_offload: return EOPNOTSUPP if rule specifies no actions

[Sasha Levin]

From: Pablo Neira Ayuso <pablo@netfilter.org>

[ Upstream commit 81ec61074bcf68acfcb2820cda3ff9d9984419c7 ]

If the rule only specifies the matching side, return EOPNOTSUPP.
Otherwise, the front-end relies on the drivers to reject this rule.

Fixes: c9626a2cbdb2 ("netfilter: nf_tables: add hardware offload support")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/netfilter/nf_tables_offload.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/netfilter/nf_tables_offload.c b/net/netfilter/nf_tables_offload.c
index 6f7eab502e65..e743f811245f 100644
--- a/net/netfilter/nf_tables_offload.c
+++ b/net/netfilter/nf_tables_offload.c
@@ -44,6 +44,9 @@ struct nft_flow_rule *nft_flow_rule_create(struct net *net,
 		expr = nft_expr_next(expr);
 	}
 
+	if (num_actions == 0)
+		return ERR_PTR(-EOPNOTSUPP);
+
 	flow = nft_flow_rule_alloc(num_actions);
 	if (!flow)
 		return ERR_PTR(-ENOMEM);
-- 
2.20.1

[PATCH AUTOSEL 5.4 038/187] netfilter: bridge: make sure to pull arp header in br_nf_forward_arp()

[Sasha Levin]

From: Eric Dumazet <edumazet@google.com>

[ Upstream commit 5604285839aaedfb23ebe297799c6e558939334d ]

syzbot is kind enough to remind us we need to call skb_may_pull()

BUG: KMSAN: uninit-value in br_nf_forward_arp+0xe61/0x1230 net/bridge/br_netfilter_hooks.c:665
CPU: 1 PID: 11631 Comm: syz-executor.1 Not tainted 5.4.0-rc8-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 <IRQ>
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x1c9/0x220 lib/dump_stack.c:118
 kmsan_report+0x128/0x220 mm/kmsan/kmsan_report.c:108
 __msan_warning+0x64/0xc0 mm/kmsan/kmsan_instr.c:245
 br_nf_forward_arp+0xe61/0x1230 net/bridge/br_netfilter_hooks.c:665
 nf_hook_entry_hookfn include/linux/netfilter.h:135 [inline]
 nf_hook_slow+0x18b/0x3f0 net/netfilter/core.c:512
 nf_hook include/linux/netfilter.h:260 [inline]
 NF_HOOK include/linux/netfilter.h:303 [inline]
 __br_forward+0x78f/0xe30 net/bridge/br_forward.c:109
 br_flood+0xef0/0xfe0 net/bridge/br_forward.c:234
 br_handle_frame_finish+0x1a77/0x1c20 net/bridge/br_input.c:162
 nf_hook_bridge_pre net/bridge/br_input.c:245 [inline]
 br_handle_frame+0xfb6/0x1eb0 net/bridge/br_input.c:348
 __netif_receive_skb_core+0x20b9/0x51a0 net/core/dev.c:4830
 __netif_receive_skb_one_core net/core/dev.c:4927 [inline]
 __netif_receive_skb net/core/dev.c:5043 [inline]
 process_backlog+0x610/0x13c0 net/core/dev.c:5874
 napi_poll net/core/dev.c:6311 [inline]
 net_rx_action+0x7a6/0x1aa0 net/core/dev.c:6379
 __do_softirq+0x4a1/0x83a kernel/softirq.c:293
 do_softirq_own_stack+0x49/0x80 arch/x86/entry/entry_64.S:1091
 </IRQ>
 do_softirq kernel/softirq.c:338 [inline]
 __local_bh_enable_ip+0x184/0x1d0 kernel/softirq.c:190
 local_bh_enable+0x36/0x40 include/linux/bottom_half.h:32
 rcu_read_unlock_bh include/linux/rcupdate.h:688 [inline]
 __dev_queue_xmit+0x38e8/0x4200 net/core/dev.c:3819
 dev_queue_xmit+0x4b/0x60 net/core/dev.c:3825
 packet_snd net/packet/af_packet.c:2959 [inline]
 packet_sendmsg+0x8234/0x9100 net/packet/af_packet.c:2984
 sock_sendmsg_nosec net/socket.c:637 [inline]
 sock_sendmsg net/socket.c:657 [inline]
 __sys_sendto+0xc44/0xc70 net/socket.c:1952
 __do_sys_sendto net/socket.c:1964 [inline]
 __se_sys_sendto+0x107/0x130 net/socket.c:1960
 __x64_sys_sendto+0x6e/0x90 net/socket.c:1960
 do_syscall_64+0xb6/0x160 arch/x86/entry/common.c:291
 entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x45a679
Code: ad b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00 00 66 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 7b b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007f0a3c9e5c78 EFLAGS: 00000246 ORIG_RAX: 000000000000002c
RAX: ffffffffffffffda RBX: 0000000000000006 RCX: 000000000045a679
RDX: 000000000000000e RSI: 0000000020000200 RDI: 0000000000000003
RBP: 000000000075bf20 R08: 00000000200000c0 R09: 0000000000000014
R10: 0000000000000000 R11: 0000000000000246 R12: 00007f0a3c9e66d4
R13: 00000000004c8ec1 R14: 00000000004dfe28 R15: 00000000ffffffff

Uninit was created at:
 kmsan_save_stack_with_flags mm/kmsan/kmsan.c:149 [inline]
 kmsan_internal_poison_shadow+0x5c/0x110 mm/kmsan/kmsan.c:132
 kmsan_slab_alloc+0x97/0x100 mm/kmsan/kmsan_hooks.c:86
 slab_alloc_node mm/slub.c:2773 [inline]
 __kmalloc_node_track_caller+0xe27/0x11a0 mm/slub.c:4381
 __kmalloc_reserve net/core/skbuff.c:141 [inline]
 __alloc_skb+0x306/0xa10 net/core/skbuff.c:209
 alloc_skb include/linux/skbuff.h:1049 [inline]
 alloc_skb_with_frags+0x18c/0xa80 net/core/skbuff.c:5662
 sock_alloc_send_pskb+0xafd/0x10a0 net/core/sock.c:2244
 packet_alloc_skb net/packet/af_packet.c:2807 [inline]
 packet_snd net/packet/af_packet.c:2902 [inline]
 packet_sendmsg+0x63a6/0x9100 net/packet/af_packet.c:2984
 sock_sendmsg_nosec net/socket.c:637 [inline]
 sock_sendmsg net/socket.c:657 [inline]
 __sys_sendto+0xc44/0xc70 net/socket.c:1952
 __do_sys_sendto net/socket.c:1964 [inline]
 __se_sys_sendto+0x107/0x130 net/socket.c:1960
 __x64_sys_sendto+0x6e/0x90 net/socket.c:1960
 do_syscall_64+0xb6/0x160 arch/x86/entry/common.c:291
 entry_SYSCALL_64_after_hwframe+0x44/0xa9

Fixes: c4e70a87d975 ("netfilter: bridge: rename br_netfilter.c to br_netfilter_hooks.c")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Reviewed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/bridge/br_netfilter_hooks.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/bridge/br_netfilter_hooks.c b/net/bridge/br_netfilter_hooks.c
index af7800103e51..59980ecfc962 100644
--- a/net/bridge/br_netfilter_hooks.c
+++ b/net/bridge/br_netfilter_hooks.c
@@ -662,6 +662,9 @@ static unsigned int br_nf_forward_arp(void *priv,
 		nf_bridge_pull_encap_header(skb);
 	}
 
+	if (unlikely(!pskb_may_pull(skb, sizeof(struct arphdr))))
+		return NF_DROP;
+
 	if (arp_hdr(skb)->ar_pln != 4) {
 		if (is_vlan_arp(skb, state->net))
 			nf_bridge_push_encap_header(skb);
-- 
2.20.1